Skip to content

AI Processors

AI Processors are specialized AI agents within NetOrca Pack that perform specific tasks across the platform. Depending on the type, an AI Processor can generate infrastructure configurations, validate change requests, verify generated outputs, optimize prompts, or render structured dynamic UI responses — all powered by LLMs configured by administrators. Service Owners define AI Processors for each of their services, choosing the appropriate type for each use case.

AI Processor Types

AI Processors are categorized by where they operate in the workflow:

  • Pack Pipeline — Three sequential stages (CONFIG, VERIFY, EXECUTION) that generate, validate, and finalize some form of configurations required in deployments. See Pack Pipeline deep dive for details.
  • Change Instance Validator — Automated validation of consumer requested declarations (i.e. Change Instance) before processing.
  • Pack Optimizer — Optimizes prompts to reduce the number of retriggers in self-healing loops. Available when the Universal Executor is enabled.

For detailed descriptions of each type, see AI Processor Types.

Setting Up AI Processors

AI Processors are configured by Service Owners after LLM Models have been set up by administrators. Each service can have one AI Processor for each stage.

Each field in the request body is explained in its dedicated documentation page — see Prompt Details, Response Schema, Generative UI, and Pack Pipeline for more details.

To set up an AI Processor:

POST /v1/external/serviceowner/ai_processors/ HTTP/1.1
Content-Type: application/json
Authorization: Token <YOUR_TOKEN>

{
  "name": "config for firewall rule",
  "service": <service_id>,
  "llm_model": <llm_model_id>,
  "action_type": "config",
  "prompt": "Generate an AWS AS3 and PaloAlto json configuration for the given service item declaration.",
  "active": false,
  "response_schema": null,
  "extra_data": {
    "include_change_instance": true,              // Pack Pipeline
    "enable_pack_context": true,                  // Pack Pipeline
    "include_previous_declaration": false,        // Pack Pipeline
    "include_previous_pipeline_data": [],         // Pack Pipeline
    "include_service_config": true,               // Pack Pipeline
    "enable_generative_ui": true,                 // Pack Pipeline
    "generative_ui_schema": null,                 // Pack Pipeline
    "schedule_enabled": false,                    // Optimizer
    "schedule_crontab": "",                       // Optimizer
    "allow_auto_approval": false,                 // Change Instance Validator
    "allow_auto_rejection": false,                // Change Instance Validator
    "send_service_info": false,                   // Change Instance Validator
    "send_existing_service_items": false          // Change Instance Validator
  }
}

Enabling and Disabling AI Processors

Some AI Processors are triggered automatically as part of the workflow:

  • In the Pack Pipeline, the VERIFY stage runs automatically after CONFIG generates its output, and EXECUTION runs automatically after VERIFY completes.
  • The Change Instance Validator is triggered automatically when a consumer submits a declaration and the resulting change instance enters a pending state.

An AI Processor can be disabled to prevent it from being triggered automatically, without removing its configuration.

PUT /v1/external/serviceowner/ai_processors/<ai_processor_id>/ HTTP/1.1
Content-Type: application/json
Authorization: Token <YOUR_TOKEN>
{
  "active": false
}